System and Method For Designing of Dictionaries For Sparse Representation

Abstract

A signal processing system adapted for sparse representation of signals is provided, comprising: (i) one or more training signals; (ii) a dictionary containing signal-atoms; (iii) a representation of each training signal using a linear combination of said dictionary's signal-atoms; (iv) means for updating the representation of the training signal; (v) means for updating the dictionary one group of atoms at a time, wherein each atom update may include all representations referring to said updated atom; and (vi) means for iterating (iv) and (v) until a stopping rule is fulfilled. The system uses the K-SVD algorithm for designing dictionaries for sparse representation of signals.

Description

FIELD OF THE INVENTION[0001]The present invention relates to a system and method for the sparse representation of signals. The invention is particularly relevant for applications such as compression, regularization in inverse problems, feature extraction, denoising, separation of texture and cartoon content in images, signal analysis, signal synthesis, inpainting and restoration.BACKGROUND OF THE INVENTIONSparse Representation of Signals[0002]In recent years there has been a growing interest in the study of sparse representation of signals. Using an overcomplete dictionary that contains prototype signal-atoms, signals are described as sparse linear combinations of these atoms. Applications that use sparse representation are many and include compression, regularization in inverse problems, feature extraction, and more. Recent activity in this field concentrated mainly on the study of pursuit algorithms that decompose signals with respect to a given dictionary. Designing dictionaries ...

AI Technical Summary

Benefits of technology

[0072]The invention thus relate to a novel system and algorithm for adapting dictionaries so as to represent signals sparsely. Given a set of training signals we seek the dictionary D that leads to the best possible representations for each member in this set with strict sparsity constraints. The invention introduces the K-SVD algorithm that addresses the above task, generalizing the K-Means algorithm. The K-SVD is an iterative method that alternates between sparse coding of the examples based on the current dictionary, and an update process for the dictionary atoms so as to better fit the data. The update of the dictionary columns is done jointly with an update of the sparse representation coefficients related to it, resulting in accelerated convergence. The K-SVD algorithm is flexible and can work with any pursuit method, thereby tailoring the dictionary to the application in mind.

[0077]The process of updating only one column of D at a time is a problem having a straightforward solution based on the singular value decomposition (SVD). Furthermore, allowing a change in the coefficients' values while updating the dictionary columns accelerates convergence, since the subsequent columns updates will be based on more relevant coefficients. The overall effect is very much in line with the leap from gradient descent to Gauss-Seidel methods in optimization.

Problems solved by technology

These dictionaries have the potential to outperform commonly used pre-determined dictionaries.

Images